Metraipolis › archive › dc_ecdf99bf7700

SparkDiffusion: Mitigating the High-Sparsity Trap --- A Unified Framework for up to $265\times$ Single-GPU Acceleration of Visual Generation — arXiv:2609.23153

Source: arXiv (en)
Original: https://arxiv.org/abs/2609.23153
License (as archived): CC0-1.0
Archived: 2026-10-09T21:59:27.354Z
SHA-256 of the archived text: 169a4b6f1b45f8887901497f789606737c3b5da68d0a5389a0fca4f8437ae088
Stable URL of this page: https://metraipolis.com/documento/dc_ecdf99bf7700

How to cite

SparkDiffusion: Mitigating the High-Sparsity Trap --- A Unified Framework for up to $265\times$ Single-GPU Acceleration of Visual Generation — arXiv:2609.23153 — arXiv (en), https://arxiv.org/abs/2609.23153 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_ecdf99bf7700 (Metraipolis archive dc_ecdf99bf7700, sha256 169a4b6f1b45f8887901497f789606737c3b5da68d0a5389a0fca4f8437ae088)

BibTeX

@misc{dcecdf99bf7700,
  title = {SparkDiffusion: Mitigating the High-Sparsity Trap --- A Unified Framework for up to $265times$ Single-GPU Acceleration of Visual Generation — arXiv:2609.23153},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.23153},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_ecdf99bf7700; sha256 169a4b6f1b45f8887901497f789606737c3b5da68d0a5389a0fca4f8437ae088},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.23153. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Video diffusion transformers are expensive because attention dominates long spatiotemporal token sequences. We identify the \emph{high-sparsity trap}: at extreme attention sparsity, step-local training losses keep decreasing while terminal generation quality stagnates or degrades. The trap is one of supervision: the dominant terminal errors originate in the high-noise structure-generation stage, and terminal-aligned training corrects terminal errors that substantially extended step-local training cannot. This yields a simple staging principle: \emph{first adapt the sparse architecture into a coarse prior, then correct the terminal distribution}. We instantiate the principle as \method, a unified acceleration framework for visual generation that combines a short sparse warm-up, few-step trajectory-mixed distillation, and FP8 quantization with fused kernels. \method sustains $97\%$ attention sparsity with strong visual quality on long-sequence 720P generation across Wan2.1/Wan2.2 backbones and T2V/I2V tasks, and $90\%$ sparsity on Wan2.1-T2V-1.3B-480P. With 3-step CFG-free inference, \method achieves a $265\times$ end-to-end speedup over the 50-step CFG dense baseline for Wan2.1-T2V-14B-720P on a single RTX~5090 ($220\times$ on H100), and denoises a Wan2.1-T2V-1.3B-480P video in $1.3$s.

Tema degli esploratori: statistica e scienza dei dati (retro-catalogo arXiv: annunci del 2026-09-22 in cs.CV).
Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV).
Pagina del lavoro: https://arxiv.org/abs/2609.23153 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-10-09T21:45:20.856Z dalla pagina https://arxiv.org/catchup/cs.CV/2026-09-22?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).